WORKLOAD QUALIFICATION FRAMEWORK FOR DATA STORES

Abstract

Data Store migration involves data transfer among data storage systems, data formats &/or computer systems. Data
Store migration is often considered a difficult subject as it requires careful adaptation of data structures & data models
that are appropriate with the target data store without affecting the functionality of application, & correctness and
integrity of data. With rapid evolution of cloud services, migrating from legacy data stores to cloud native, or open-source
data stores is gaining wide attention. However, there is no standardized approach which addresses the problem of
workload categorization, & simplifying the choice of target data stores. This paper presents a comprehensive workload
qualification framework & shares an example of choosing between relational & non-relational (key value) data stores.
The paper addresses the pervasive challenges and lack of standardization in migrating traditional data stores to open
source, cloud native data stores by proposing a unified framework. This framework encompasses algorithm migration,
model migration, and migration schemes, tailored to enhance the data migration process across various environments.
Through comparative analysis and integration of existing frameworks, it aims to offer a robust solution that aids
organizations and developers in navigating the complexities of data migration, thereby contributing to more efficient and
effective database management and transition strategies